QUANTITATIVE PREDICTION AND PRIORITIZATION FOR REFORESTATION SUCCESS: DEVELOPMENT OF AN AUTOMATED DECISION SUPPORT TOOL FOR THE USDA FOREST SERVICE

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: ACTIVE
• Funding Source: AFRI COMPETITIVE GRANT
• Reporting Frequency: Annual
• Accession No.: 1027720
• Grant No.: 2022-67019-36438
• Cumulative Award Amt.: $649,548.00
• Proposal No.: 2021-09817
• Project Start Date: Nov 1, 2021
• Project End Date: Oct 31, 2025
• Grant Year: 2022
• Program Code: [A1451]- Renewable Energy, Natural Resources, and Environment: Agroecosystem Management

Recipient Organization
UNIVERSITY OF MONTANA
COLLEGE OF FORESTRY AND CONSERVATION
MISSOULA,MT 59812

Performing Department
Geosciences

Non Technical Summary
Forests in the semiarid western US provide key ecosystem services for the region and are critical for local economies and populations. However, risks to forests are accelerating in the US and globally. Resource managers are seeing a dramatic increase in reforestation needs in recent decades driven by the combination of more burned area, the accelerated pace of timber harvest, and the US engagement in the trillion tree initiative. This is presenting a challenge for the USFS, which has a legal mandate to maintain the extent of national forest land and to monitor forest recovery following disturbance or harvest. This project presents a comprehensive methodology that will allow the USFS and other agencies to prioritize planting efforts towards areas where tree planting is needed for forest recovery while minimizing expenditures towards areas that will no longer support trees or areas that will likely regenerate naturally. Our approach is based on an ensemble of existing models that will determine the probability of seed availability, the probability of suitable hydroclimatic conditions for seedling establishment, and the intersection of seed availability and suitable hydroclimate which defines areas where natural regeneration is likely to occur. This information is used within a risk-benefit framework to inform resource allocation and to recommend management goals better aligned with the new realities of western forests.

Animal Health Component

60%

Research Effort Categories

Basic

10%

Applied

60%

Developmental

30%

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
123	0612	1070	80%
122	0612	2050	20%

Knowledge Area
122 - Management and Control of Forest and Range Fires; 123 - Management and Sustainability of Forest Resources;

Subject Of Investigation
0612 - Conifer forests of the West;

Field Of Science
2050 - Hydrology; 1070 - Ecology;

Keywords

decision support systems

forest management

forestry

reforestation

resource optimization

Goals / Objectives
The overarching goalof this project is to develop a web-based decision support system (DSS) that will allow the USFS and other federal, state, and local agencies to prioritize planting efforts towards areas where tree planting is needed for forest recovery while minimizing expenditures towards areas that will no longer support trees or areas that will likely regenerate naturally. Our approach results in a fully automated tool thatuses national data layers along with quantitative information on tree regeneration derived from peer-reviewed research to overcome the limitations of current tools used by forest managers at the US Forest Service. The specific objectives of this project to achieve the overarching goals are:1) We will combine an existing physically-based solution ofsoil moisture and soil surface temperature dynamics resolved at 100 m resolution and 4 hourly time steps at thousands of strategic key locations using geostatistical interpolation methods to produce high-resolution grids of soil moisture and temperature over forested regions of interest to our USFS partners.2) We will use the soil moisture and temperature grids developed in objective 1) to run ourecohydrologicmodels, calculate seedling thermal and hydraulic stress and quantify climate suitability for regeneration at burned areas targeted for potential reforestation efforts.3) We will use combine ourextensive body of empirical research and published meta-analyses of conspecific basal area as a proxy of the abundance of seed sources (adult trees) to quantify seed availability at burned areas targeted for potential reforestation efforts.4) We will evaluate the quality of model predictions (objectives 1 and 2) andformally assess model predictive uncertainty.5) We will combine products from objectives 2), 3) and 4) along withancillary user inputs to develop a risk-based cost-benefit analysis of forest regeneration effort that balances investment, simulated probability of regeneration success, model uncertainty, and risk of regeneration failure.6) We will implement simulated climate suitability for regeneration, seed availability quantification,and our cost-benefit analysis tools within an interactive web-based DSS accessible to partnering USFS forest managers.7) We will actively work with our USFS partners to disseminate and facilitate the adoption of our DSS within the agency.

Project Methods
The main outcome of this project is a web-based decision support system (DSS) that will allow the USFS 1) to prioritize planting efforts towards areas where tree planting is needed for forest recovery while minimizing expenditures towards areas that will no longer support trees or areas that will likely regenerate naturally, and 2) select adequate management strategies that are better aligned with the changing environmental conditions of lowertreelineforests in the western US. The DSS uses an adaptation of mechanistic and statistical models previously developed by our team to evaluate a site's climate suitability for seedling recruitment and the likelihood of abundant seeds after awildfire. The method of site triage for management action rests on the idea that burned sites with a climate suitable for seedling survival and abundant seeds will recover without action, whileat sites that no longer have a climate suitable for seedling survival the risk of reforestation failure is too high to justify investment. Other factors being equal, these sites should receive lower priority for resource allocation and should be managed following new adaptive management strategies.Thescientific and technical results of the project will be analyzed and evaluated using a combination of data collected byour USFS partners and remote sensing data. At individual sites, we will evaluate the quality of our seedling mortality model against field surveys. The USFS monitors seedling survival at sites where seedlings are out-planted using marked rows ("stake rows") of at least 25 individual trees. Survival at each stake row is then assessed at years 1 and 3 following planting.These data are collected across all western U.S. forests and will be made available to our team by the USFS Region 1 Director of Forest Management. Data on the survival of out-planted seedlings, collected at more than 10,000 sites across multiple years, provide an opportunity to examine the effects of drought and heat stress on the survival of conifer seedlings across large domains.At the landscape scale, we will compare our model projections of survival to the current distribution of forest cover near the lowertreeline. We will use two datasets; the250-meterresolution MODIS Vegetation Continuous Fields (VCF) product and the 30 m LANDFIRE canopy height product to represent current forest cover.Project progress, model quality, and milestones will be further assessed by our USFS partners during planned annual stakeholder workshops. Resources are budgeted for these annual events. During years 3 and 4 we will start discussing in these meetings potential options to facilitate the integration of the new methodology within the operational protocols of the USFS Region 1 and 4 and how to reach out to other regions.In a preliminary meeting with our USFS partners, we have discussed a number of quantitative and qualitative metrics that can be used to assess the project impacts in terms of the projected outcomes. The baseline conditions are given by the application of the current Decision Support Tool (DST0) available to USFS Region 1 and 4. These specific metrics we will use to gauge success:1) Number of reforestation sites where DSS is applied2) Number of USFS Forest Regions that adopt or are interested in adopting our proposed DSS3) Skill of our DSS to predict mortality in active reforestation sites based onstakerowdata4) Skill of our DSS to predict natural recovery in disturbed sites based on RS and surveys5) Stakeholder assessment of proposed site prioritization from risk benefit analysis framework6) Stakeholder assessment of adaptiverecommendation strategy maps generated by our DSS

Progress 11/01/23 to 10/31/24

Outputs
Target Audience: During the project's third year, the primary audience was forest managers at US Forest Service, and secondarily other natural resource managers at other organizations. During this period, we have met with partners to review project goals, receive feedback on our decision support system (RegenMapper, partially developed in a previous project and currently adopted as the primary tool for reforestation assessments), and plan the necessary advances and improvements in this tool to help USFS forest managers prioritize planting efforts in a cost-effective manner, as outlined in the overarching goals of the project. CoPI Holden has met with multiple USFS stakeholder groups to introduce the tool and promote its use in the following dissemination efforts: February 21, 2024: Region1- Region 4 Reforestation/SI Workshop November 20th 2024. workshop for Bitterroot NF. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided? This reporting period we have provided partial or full support and mentorship to 1 graduate student(Lewis Faller, degree expected in 2025), that contributed to advancing the goals of this project. Mentoring activities include periodic meetings to discuss technical and scientific issues, provide feedbacks on their analysis and technical writing, and other standard mentoring practices. How have the results been disseminated to communities of interest? As prevously outlined we met with project stakeholders where we reviewed progress, project assessment metrics, and received feedback on the DSS. In addition CoPI Holden has met consistently with multiple USFS stakeholder groups during the performance period. This outreach included: February 21, 2024: Region1- Region 4 Reforestation/SI Workshop November 20th 2024. workshop for Bitterroot NF. What do you plan to do during the next reporting period to accomplish the goals?For the final year of the project we expect the Masters student (Faller) to complete his MS degree,publish his work on spatial optimiazation of planting efforts, and incorporate FORSYS outputs into the online decision support tool.

Impacts
What was accomplished under these goals? 1. We have developed a model that simulates climate suitability for regeneration, seed availability, and cost associated with distance to roads which is a key driver of the cost of reforestation projects. We have embedded these models within aweb-based DSS accessible to partnering USFS forest managers. We published two articles this reporting period. The first describes a kybiophysical variable we simulated (Potential land surface temperature) and describes how PSST constrains the distribution of natural regneration in forests of the western US (Holden et al. 2024). The second paper (Dobrowski et al. 2024) examines the reforestation pipeline in the western US focusing on both supply and demand constraints. 2) We have Integratedthe climate suitability, seed availability, and cost surfaces within a spatial optimization platform (FORSYS). FORSYS has been used by the USFS to spatially optimize where to conduce wilfire fuels reduction treatments. We have used the platform to identify optimal planting patterns for USFS end users within fire perimeters. 2. Disseminate and facilitate the adoption of our DSS within the USFS A priority during this performance period was to engage with stakeholders in the forest service to review the project goals, establish scientific and technical priorities, receive feedback, and promote the use Regenmapper to inform decision-making within the agency. CoPI Holden meets often with stakeholders at the USFS and acts as a communication bridge between our team and the end users. He has also engaged in formal outreach events, including a presentation our DSS prototype in multiple workshops, and during demonstrations with silviculturalists.

Publications

• Type: Other Journal Articles Status: Published Year Published: 2024 Citation: Holden, Z.A., Dobrowski, S.Z., Swanson, A. et al. Low-elevation forest extent in the western United States constrained by soil surface temperatures. Nat. Geosci. 17, 12491253 (2024). https://doi.org/10.1038/s41561-024-01577-0
• Type: Other Journal Articles Status: Published Year Published: 2024 Citation: Dobrowski SZ, Aghai MM, Chichilnisky du Lac A, Downer R, Fargione J, Haase DL, Hoecker T, Kildisheva OA, Murdoch A, Newman S, North M, Saksa P, Sjoholm M, Baribault T, Buonanduci MS, Chambers ME, Gonzales-Kramer L, Harvey BJ, Hurteau MD, Loevner J, Safford HD and Sloan J (2024) Mind the Gapreforestation needs vs. reforestation capacity in the western United States. Front. For. Glob. Change 7:1402124. doi: 10.3389/ffgc.2024.1402124

Progress 11/01/22 to 10/31/23

Outputs
Target Audience:During the project's second year, the primary audience was forest managers at US Forest Service, and secondarily other natural resource managers at other organizations. During this period, we have met with partners to review project goals, receive feedback on our decision support system (RegenMapper, partially developed in a previous project and currently adopted as the primary tool for reforestation assessments), and plan the necessary advances and improvements in this tool to help USFS forest managers prioritize planting efforts in a cost-effective manner, as outlined in the overarching goals of the project. CoPI Holden has met with multiple USFS stakeholder groups to introduce the tool and promote its use in the following dissemination efforts: Region 1 - multiple 45 minute presentation focused on Regenmapper. Region 1 - 10/23, reforestation implementation workshop. Missoula, MT. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?This reporting period we have provided partial or full support and mentorship to 2 graduate students (Mr. Michael Kendree, MSc, degree expected in 2024; Mr. Drew Lyons MS, degree expected in 2025), that contributed to advancing the goals of this project. Mentoring activities include periodic meetings to discuss technical and scientific issues, provide feedbacks on their analysis and technical writing, and other standard mentoring practices. How have the results been disseminated to communities of interest?As prevously outlined we met with project stakeholders where we reviewed progress, project assessment metrics, and received feedback on the DSS. In addition CoPI Holden has met consistently with multiple USFS stakeholder groups during the performance period. This outreach included: Region 1 Soil/Water/Fish meeting in Missoula. 45 minute presentation focused on Regenmapper. Updated regenmapper documentation Region 1 and Washington Office reforestation implementation workshop. Missoula, MT. What do you plan to do during the next reporting period to accomplish the goals?We will contine to address specific goals and communicate this work via peer reviewed publications. We have developed key capacity to incorporate a cost-benefit framework in RegenMapper which is currently in initial design and development.

Impacts
What was accomplished under these goals? The project is organized in nine work packages with technical tasks that advance seven specific goals: Combine an physically-based solutions of soil moisture and soil surface temperature dynamics resolved at 100 m resolution and 4 hourly time steps at thousands of strategic key locations We have completed the development and processing of gridded climate time series data, weather station time series, and stream gauge station time series, soil moisture time series, and gridded soils data, needed to execute and calibrate a ecohydrologic model that simulates plant stress. Accurate time series of Soil moisture observations are critical for model calibration and verification. This year we completed a collection of soil moisture observation for 665 sites across the western US. Use the soil moisture and temperature grids developed in objective to run our ecohydrologic models, calculate seedling thermal and hydraulic stress, and predict seedling mortality We have made notable progress improving the models that will generate these two key biophysical predictors of mortality. A MS student is currently testing these improvements, that are based on a solution of the equation that govern the bulk flow of water within the vascular system of plants. This model will be run at scale with grids produced during goal 1). A MS defense and a publication resulting from this work is expected during 2023 or early 2024. Develop a model that uses conspecific basal area as a proxy of the abundance of seed sources (adult trees) to quantify seed availability at burned areas targeted for potential reforestation efforts. We used data collected and collated in previous research to develop empirical models that relate conspecific basal area to seed source availability. This has been integrated into sub-routines currently available in Regenmapper. This work has been organized and conducted by a Masters student funded by this research. We will evaluate the quality of model predictions (objectives 1 and 2) and formally assess model predictive uncertainty. Activity on the work packages was initiated in this performance period. We will combine products from objectives 2), 3) and 4) along with ancillary user inputs to develop a risk-based cost-benefit analysis of forest regeneration effort that balances investment, simulated probability of regeneration success, model uncertainty, and risk of regeneration failure. Activity on the work packages associated with this goal was initiated in this performance period. We will implement simulated climate suitability for regeneration, seed availability quantification, and our cost-benefit analysis tools within an interactive web-based DSS accessible to partnering USFS forest managers. Activity on the work packages associated with this goal was initated in this performance period. Disseminate and facilitate the adoption of our DSS within the USFS A priority during this performance period was to engage with stakeholders in the forest service to review the project goals, establish scientific and technical priorities, receive feedback, and promote the use Regenmapper to inform decision-making within the agency. CoPI Holden meets often with stakeholders at the USFS and acts as a communication bridge between our team and the end users. He has also engaged in formal outreach events, including a presentation our DSS prototype in multiple workshops, and during demonstrations with silviculturalists.

Publications

Progress 11/01/21 to 10/31/22

Outputs
Target Audience: Target audiences: During the project's first year, the primary audience was forest managers at US Forest Service, and secondarily other natural resource managers at other organizations. During this period, we have met multiple times with our partners to review project goals, receive feedback on our decision support system (RegenMapper, partially developed in a previous project and currently adopted as the primary tool for reforestation assessments), and plan the necessary advances and improvements in this tool to help USFS forest managers prioritize planting efforts in a cost-effective manner, as outlined in the overarching goals of the project. During the initial stage of the project, CoPI Holden has met with multiple USFS stakeholder groups to introduce the tool and promote its use in the following dissemination efforts: Twoformal workshops on the use of Regenmapper (see other products) Several informal tests/demos of Regenmapper for USFS silviculturalists from USFS Regions 1, 2, 4 and 5 One presentation to a regional tree seedling nursery Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Duringthis reporting period we have provided partial or full support and mentorship to one graduate student(Mr. Michael Kendree, MSc, degree expected in 2023), oneundergraduate student (Mr. Adam Viray, UG student), and onetechnician (Mr. Drew Lyons) that contributed to advancing the goals of this project. Mentoring activities include periodic meetings with them to discuss technical and scientific issues, provide feedbackon their technical writing, and other standard mentoring practices. Mr. Adam Viray completed a capstone presentation on the physics of water flow within plants and graduated with a degree in physics at the end of the Fall semester of 2022. How have the results been disseminated to communities of interest?We met virtually with project stakeholders at the USFS in the Spring of 2022to review project goalsand received feedback on the DSS. In addition to this initial meeting, and as outlined in the target audience section, CoPI Holden has met consistently with multiple USFS stakeholder groups this past year and has been active in outreach including formal presentationsat two workshops (listed under other products), conducting demos of our DSS prototypewith USFS forest managers and silviculturalists from multiple USFS Regions, and presenting to managers at a tree nursery in Idaho. What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? Theoverarching goalof this project is to model the likelihood of climate-induced tree recruitment failure in semi-arid forestsof the western USA, which is a precursor to the loss of forest coverage. This biophysical metric will be used within a risk-benefit framework to helpUSFS and other federal, state, and local agencies to prioritize more targeted and efficient planting efforts.These products will be integrated within Regenmapper, a reforestation decision support system developed by our team that is currentlyused by the USFS. The project will increase our understanding of the vulnerability of western US forests to changing climate and the extent of forests at risk of transitioning to shrubland and will encapsulate this new understanding within tools that can inform decision-making. Eventually, this project will translate into an increase in forest area recovered by reforestation efforts and a more efficient allocation of USFS reforestation resources. The project is organized into ninework packages with technical tasks that advance seven specific goals: Combine a physically-based solution ofsoil moisture and soil surface temperature dynamics resolved at 100 m resolution and 4 hourly time steps at thousands of strategic key locations We have completed the development and processing of gridded climate time series data, weather station time series, stream gauge station time series, soil moisture time series, and gridded soils data needed to execute and calibrate an ecohydrologic model that simulates plant stress.Accurate time series of Soil moisture observations are critical for model calibration and verification. This year we completed a collection of soil moisture observations for 665 sites across the western US. Use soil moisture and temperature grids developed in goal 1.to run ourecohydrologicmodels, calculate seedling thermal and hydraulic stress, and predict seedling mortality We have made notable progress in improving the models that will generate these two key biophysical predictors of mortality. AMSc student is currently testing these improvements,based on a solution of the equations that describe the bulk flow of water within the vascular system of plants. This model will be run at scale with grids produced during goal 1). A MScdefense and a publication resulting from this work are expected during 2023 or early 2024. Develop a model that uses conspecific basal area estimates as a proxy of the abundance of seed sources (adult trees) to quantify seed availability at burned areas targeted for potential reforestation efforts. We will use metanalysis and results conducted in the course of a previous project and work done by collaborators to develop this model. Activity on the work packages associated with this goal has not started yet. Evaluatethe quality of model predictions (objectives 1 and 2) andformally assess model predictive uncertainty. Activity on the work packages associated with this goal has not started yet. Combine products from objectives 2., 3.and 4.along withancillary user inputs to develop a risk-based cost-benefit analysis of forest regeneration. Activity on the work packages associated with this goal has not started yet. Implement simulated climate suitability for regeneration, seed availability quantification,and our cost-benefit analysis tools within Regenmapper, an interactive web-based DSS accessible to partnering USFS forest managers. Activity on the work packages associated with this goal has not started yet. Disseminate and facilitate the adoption of our DSS within the USFS A priority during the first year of the project was to engage with stakeholders in the forest service to review the project goals, establish scientific and technical priorities, receive feedback, and promote the use of Regenmapper to inform decision-making within the agency. CoPI Holden has metoften with stakeholders at the USFS during this past year and is actingas a communication bridge between our team and end-user partners. He has also engaged in formal outreach events, including a presentation of our DSS prototype in two separate formal workshops, and during several informal demonstrations with silviculturalists and forest managers.

Publications